(1) To develop animal, especially nonhuman primate, models that mimic human disease: We have established nonhuman primate (NHPs) models using Cynomolgus macaques for several influenza A viruses. Seasonal H1N1 & H3N2 viruses lead to either asymptomatic or fairly mild disease in Cynomolgus macaques. Gross pathology is limited to a few areas in the lungs. Histopathological investigations revealed alveolar edema and fibrin, hyaline membrane formation and type II alveolar pneumocyte hyperplasia. Animals recover quickly from the infection and clear virus within the first few days. There is limited evidence for virus shedding. (Brining 2010; Safronetz 2011) The pandemic H2N2 strain A/Singapore/1/57 leads to a moderate infection compared to the seasonal strains. Lung infiltrates, gross pathology and histopathology are in general slightly enhanced. Animals clear the infection with a delay and fully recover. There is limited evidence for virus shedding (Richt, 2012). We observed different degrees of clinical severity, gross pathology, lung infiltration and histology in macaques infected with three distinct pandemic H1N1 isolates. A Mexican strain isolated from a moderately sick human A/Mexico/4108/2009 (H1N1) (M4108) behaved similarly to the pandemic H2N2 strain. A California isolate from a sick boy with moderate disease A/California/4/2009 (H1N1) (Ca04) was more severe and another Mexico strain from a cluster of severe cases A/Mexico/4487/2009 (H1N1) (M4487) was severe in clinical disease and pathology. This clearly indicates that a variety of strains co-circulated with different pathogenic potential. (Brining 2010; Safronetz 2011; Feldmann, in preparation) Infection with the swine H2N3 virus A/swine/MS/06 was also more severe and was similar to the disease caused by Ca04. Thus, this newly described swine virus has potential to infect primates and thus, could potentially also infect humans (Richt, 2012). Furthermore, we have established the ferret model for influenza viruses. We have compared disease progression and virus replication of two pandemic H1N1 viruses (M4108, M4487). In contrast to the differences we observed in the nonhuman primate model, both infections did not really differ in the ferret model with both viruses causing mild respiratory disease. (Tsuda, in preparation) (2) To identify and characterize determinants of pathogenicity in animal models: To better understand the features that define highly lethal influenza viruses, we exploited a technology for the artificial generation and modification of influenza viruses (reverse genetics system). Using this technology, we have generated a consensus pandemic H1N1 virus (parent virus). In order to study virulence of the different strains and the mechanisms of pathogenesis we have generated selected recombinant viruses and characterized those mutants in vitro. For this we have compared the genome sequences of two early circulating pandemic H1N1 strains with different pathogenicity in humans (M4108, M4487). Sequence comparison revealed amino acid changes in 5 genome segments, PB2, PA, HA, NP and M. The recombinant viruses were generated on the background of M4108 by replacing the corresponding genomic segment. We have characterized the recombinant viruses in vitro and in vivo (ferret) and found preliminary evidence for a role of PB2 and PA. (Tsuda, in preparation) Severe influenza virus infections are often associated with bacterial co-infections. This seems also be the case with the latest pandemic H1N1 virus. In order to study a potentiating effect of influenza and bacterial co-infection we performed a co-infection study in cynomolgus macaques using a moderately severe pandemic H1N1 strain (Ca04) and Methicillin-resistant Staphylococcus aureus (MRSA). Animals infected with MRSA only were largely asymptomatic, whereas animals infected with Ca04 only developed moderate pulmonary disease. Interestingly, animals initially infected with MRSA followed by Ca04 showed a dramatic reduction in clinical signs, whereas those initially infected with Ca04 followed by MRSA showed enhanced clinical disease (Kobayashi, in preparation). We are currently working on the potential mechanism. Similar studies were performed with a seasonal viruses (H1N1, H3N2) and MRSA in which we did not see disease reduction or enhancement. (3) To identify potential targets for intervention: The inherent virulence properties of past (i.e. 1918 virus) and potential future pandemic influenza viruses (i.e. highly virulent H5N1) indicates the need for evaluating antiviral options. Therefore, the recently developed Cynomolgus macaque model for the 1918 influenza virus was used for a drug efficacy study. We demonstrated that oseltamivir phosphate is effective in preventing severe disease in Cynomolgus macaques caused by the 1918 influenza virus if given prophylactically. Efficacy was reduced in a treatment regime through emergence of oseltamivir-resistant mutants that led to the death of one of four animals. This emphasizes the importance of implementing combination therapy and vaccination strategies early in a pandemic. (Feldmann, in revision) In a first attempt to identify new targets for intervention we have performed a comparative transcriptomic analysis of acute host responses to infection with pandemic H1N1 (Ca04) in mice, macaques and swine. We observed differences in inflammatory molecules elicited, and the kinetics of their gene expression changes across these species, including genes regulated by nuclear receptor signaling molecules controlling lipid and metabolic processes impacting inflammatory responses to pandemic H1N1 influenza virus. While the heterodimeric retinoid X receptor (RXR) signaling pathway was differentially regulated during infection in each species, divergent expression of lipid-binding molecules and NFB molecules was observed across species. By comparing transcriptional changes across hosts in the context of clinical and virological measures and integrating evidence from transcription factor analyses, we were able to examine similarities and differences in acute host responses to pandemic H1N1 influenza infection across hosts. (Go, in revision) (4) To develop cross-protective vaccines and test their efficacy in the developed animal models: In collaboration (Dr. J. Rose) we have generated and characterized VSV-based vaccines that express the HK/156 (clade 0) H5 HA from the first position of the VSV genome. These vectors induce broadly cross-neutralizing antibodies against homologous and heterologous H5N1 viruses of different clades in mice. The vaccines provide complete protection against morbidity and mortality after heterologous challenge with clade 0 and clade 1 strains in animals even 1 year after vaccination. Therefore, VSV-based avian influenza vaccines are potent, broadly cross-protective pandemic vaccine candidates (Schwartz 2010). More recently we started a collaborative project (Dr. M. Jarvis) to develop a universal vaccine against influenza A viruses. Our approach is based on two observations: i) highly conserved B cell epitopes are present within two separate helical regions within the hemagglutinin stalk region, and antibodies against them affords heterosubtypic binding and protection, and ii) removal of hemagglutinin globular region can increase antibody responses against other normally poorly antigenic epitopes. We used the Cytomegalovirus vector platform for these studies, which is able to induce long-lasting immune responses (both T cell and antibody) following a single vaccination. The vectors have been made and we are currently testing immunogenicity.